[WebAssembly] Add unimplemented-simd128 subtarget feature

[tlively]

Summary

This is a third attempt, but this time we have vetted it on Windows
first. The previous errors were due to an uninitialized class member.

[aheejin]

lib/Target/WebAssembly/WebAssemblyInstrInfo.td

23

Change this to HasUnimplemented128 for consistency?

[aheejin]

lib/Target/WebAssembly/WebAssemblyInstrInfo.td

23

Oh, I meant HasUnimplementedSIMD128

lib/Target/WebAssembly/WebAssembly.td

 //===----------------------------------------------------------------------===//
 
 include "llvm/Target/Target.td"
 
 //===----------------------------------------------------------------------===//
 // WebAssembly Subtarget features.
 //===----------------------------------------------------------------------===//
 
-def FeatureSIMD128 : SubtargetFeature<"simd128", "HasSIMD128", "true",
+def FeatureSIMD128 : SubtargetFeature<"simd128", "SIMDLevel", "SIMD128",
                                       "Enable 128-bit SIMD">;
+
+def FeatureUnimplementedSIMD128 :
+      SubtargetFeature<"unimplemented-simd128",
+                       "SIMDLevel", "UnimplementedSIMD128",
+                       "Enable 128-bit SIMD not yet implemented in engines",
+                       [FeatureSIMD128]>;
+
 def FeatureAtomics : SubtargetFeature<"atomics", "HasAtomics", "true",
                                       "Enable Atomics">;
 def FeatureNontrappingFPToInt :
       SubtargetFeature<"nontrapping-fptoint",
                        "HasNontrappingFPToInt", "true",
                        "Enable non-trapping float-to-int conversion operators">;
 
 def FeatureSignExt :

lib/Target/WebAssembly/WebAssemblyFastISel.cpp

 #include "llvm/IR/Operator.h"
 #include "llvm/IR/PatternMatch.h"
 
 using namespace llvm;
 using namespace PatternMatch;
 
 #define DEBUG_TYPE "wasm-fastisel"
 
-extern cl::opt<bool> EnableUnimplementedWasmSIMDInstrs;
-
 namespace {
 
 class WebAssemblyFastISel final : public FastISel {
   // All possible address modes.
   class Address {
   public:
     typedef enum { RegBase, FrameIndexBase } BaseKind;
 
MVT::SimpleValueType getLegalType(MVT::SimpleValueType VT) {
     case MVT::v8i16:
     case MVT::v4i32:
     case MVT::v4f32:
       if (Subtarget->hasSIMD128())
         return VT;
       break;
     case MVT::v2i64:
     case MVT::v2f64:
-      if (Subtarget->hasSIMD128() && EnableUnimplementedWasmSIMDInstrs)
+      if (Subtarget->hasUnimplementedSIMD128())
         return VT;
       break;
     default:
       break;
     }
     return MVT::INVALID_SIMPLE_VALUE_TYPE;
   }
   bool computeAddress(const Value *Obj, Address &Addr);

lib/Target/WebAssembly/WebAssemblyISelDAGToDAG.cpp

 #include "llvm/Support/Debug.h"
 #include "llvm/Support/KnownBits.h"
 #include "llvm/Support/MathExtras.h"
 #include "llvm/Support/raw_ostream.h"
 using namespace llvm;
 
 #define DEBUG_TYPE "wasm-isel"
 
-extern cl::opt<bool> EnableUnimplementedWasmSIMDInstrs;
-
 //===--------------------------------------------------------------------===//
 /// WebAssembly-specific code to select WebAssembly machine instructions for
 /// SelectionDAG operations.
 ///
 namespace {
 class WebAssemblyDAGToDAGISel final : public SelectionDAGISel {
   /// Keep a pointer to the WebAssemblySubtarget around so that we can make the
   /// right decision when generating code for different targets.

lib/Target/WebAssembly/WebAssemblyISelLowering.h

private:
   SDValue LowerGlobalAddress(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerExternalSymbol(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerBR_JT(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerJumpTable(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVASTART(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerCopyToReg(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerINTRINSIC_VOID(SDValue Op, SelectionDAG &DAG) const;
+  SDValue LowerSIGN_EXTEND_INREG(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerAccessVectorElement(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerShift(SDValue Op, SelectionDAG &DAG) const;
 };
 
 namespace WebAssembly {
 FastISel *createFastISel(FunctionLoweringInfo &funcInfo,
                          const TargetLibraryInfo *libInfo);
 } // end namespace WebAssembly
 
 } // end namespace llvm
 
 #endif

lib/Target/WebAssembly/WebAssemblyISelLowering.cpp

 #include "llvm/Support/Debug.h"
 #include "llvm/Support/ErrorHandling.h"
 #include "llvm/Support/raw_ostream.h"
 #include "llvm/Target/TargetOptions.h"
 using namespace llvm;
 
 #define DEBUG_TYPE "wasm-lower"
 
-// Emit proposed instructions that may not have been implemented in engines
-cl::opt<bool> EnableUnimplementedWasmSIMDInstrs(
-    "wasm-enable-unimplemented-simd",
-    cl::desc("Emit potentially-unimplemented WebAssembly SIMD instructions"),
-    cl::init(false));
-
 WebAssemblyTargetLowering::WebAssemblyTargetLowering(
     const TargetMachine &TM, const WebAssemblySubtarget &STI)
     : TargetLowering(TM), Subtarget(&STI) {
   auto MVTPtr = Subtarget->hasAddr64() ? MVT::i64 : MVT::i32;
 
   // Booleans always contain 0 or 1.
   setBooleanContents(ZeroOrOneBooleanContent);
   // Except in SIMD vectors
WebAssemblyTargetLowering::WebAssemblyTargetLowering(
   addRegisterClass(MVT::i64, &WebAssembly::I64RegClass);
   addRegisterClass(MVT::f32, &WebAssembly::F32RegClass);
   addRegisterClass(MVT::f64, &WebAssembly::F64RegClass);
   if (Subtarget->hasSIMD128()) {
     addRegisterClass(MVT::v16i8, &WebAssembly::V128RegClass);
     addRegisterClass(MVT::v8i16, &WebAssembly::V128RegClass);
     addRegisterClass(MVT::v4i32, &WebAssembly::V128RegClass);
     addRegisterClass(MVT::v4f32, &WebAssembly::V128RegClass);
-    if (EnableUnimplementedWasmSIMDInstrs) {
+    if (Subtarget->hasUnimplementedSIMD128()) {
       addRegisterClass(MVT::v2i64, &WebAssembly::V128RegClass);
       addRegisterClass(MVT::v2f64, &WebAssembly::V128RegClass);
     }
   }
   // Compute derived properties from the register classes.
   computeRegisterProperties(Subtarget->getRegisterInfo());
 
   setOperationAction(ISD::GlobalAddress, MVTPtr, Custom);
for (auto Op :
         ISD::SRL_PARTS, ISD::ADDC, ISD::ADDE, ISD::SUBC, ISD::SUBE}) {
     for (auto T : {MVT::i32, MVT::i64}) {
       setOperationAction(Op, T, Expand);
     }
     if (Subtarget->hasSIMD128()) {
       for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32}) {
         setOperationAction(Op, T, Expand);
       }
-      if (EnableUnimplementedWasmSIMDInstrs) {
+      if (Subtarget->hasUnimplementedSIMD128()) {
         setOperationAction(Op, MVT::v2i64, Expand);
       }
     }
   }
 
   // There is no i64x2.mul instruction
   setOperationAction(ISD::MUL, MVT::v2i64, Expand);
 
   // We have custom shuffle lowering to expose the shuffle mask
   if (Subtarget->hasSIMD128()) {
     for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v4f32}) {
       setOperationAction(ISD::VECTOR_SHUFFLE, T, Custom);
     }
-    if (EnableUnimplementedWasmSIMDInstrs) {
+    if (Subtarget->hasUnimplementedSIMD128()) {
       setOperationAction(ISD::VECTOR_SHUFFLE, MVT::v2i64, Custom);
       setOperationAction(ISD::VECTOR_SHUFFLE, MVT::v2f64, Custom);
     }
   }
 
   // Custom lowering since wasm shifts must have a scalar shift amount
   if (Subtarget->hasSIMD128()) {
     for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32})
       for (auto Op : {ISD::SHL, ISD::SRA, ISD::SRL})
         setOperationAction(Op, T, Custom);
-    if (EnableUnimplementedWasmSIMDInstrs)
+    if (Subtarget->hasUnimplementedSIMD128())
       for (auto Op : {ISD::SHL, ISD::SRA, ISD::SRL})
         setOperationAction(Op, MVT::v2i64, Custom);
   }
 
   // There are no select instructions for vectors
   if (Subtarget->hasSIMD128())
     for (auto Op : {ISD::VSELECT, ISD::SELECT_CC, ISD::SELECT}) {
       for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v4f32})
         setOperationAction(Op, T, Expand);
-      if (EnableUnimplementedWasmSIMDInstrs)
+      if (Subtarget->hasUnimplementedSIMD128())
         for (auto T : {MVT::v2i64, MVT::v2f64})
           setOperationAction(Op, T, Expand);
     }
 
   // As a special case, these operators use the type to mean the type to
   // sign-extend from.
   setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i1, Expand);
   if (!Subtarget->hasSignExt()) {
+    // Sign extends are legal only when extending a vector extract
+    auto Action = Subtarget->hasSIMD128() ? Custom : Expand;
     for (auto T : {MVT::i8, MVT::i16, MVT::i32})
-      setOperationAction(ISD::SIGN_EXTEND_INREG, T, Expand);
+      setOperationAction(ISD::SIGN_EXTEND_INREG, T, Action);
   }
   for (auto T : MVT::integer_vector_valuetypes())
     setOperationAction(ISD::SIGN_EXTEND_INREG, T, Expand);
 
   // Dynamic stack allocation: use the default expansion.
   setOperationAction(ISD::STACKSAVE, MVT::Other, Expand);
   setOperationAction(ISD::STACKRESTORE, MVT::Other, Expand);
   setOperationAction(ISD::DYNAMIC_STACKALLOC, MVTPtr, Expand);
for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v2i64, MVT::v4f32,
           for (auto Ext : {ISD::EXTLOAD, ISD::ZEXTLOAD, ISD::SEXTLOAD})
             setLoadExtAction(Ext, T, MemT, Expand);
         }
       }
     }
   }
 
   // Expand additional SIMD ops that V8 hasn't implemented yet
-  if (Subtarget->hasSIMD128() && !EnableUnimplementedWasmSIMDInstrs) {
+  if (Subtarget->hasSIMD128() && !Subtarget->hasUnimplementedSIMD128()) {
     setOperationAction(ISD::FSQRT, MVT::v4f32, Expand);
     setOperationAction(ISD::FDIV, MVT::v4f32, Expand);
   }
 
   // Custom lower lane accesses to expand out variable indices
   if (Subtarget->hasSIMD128()) {
     for (auto T : {MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v4f32}) {
       setOperationAction(ISD::EXTRACT_VECTOR_ELT, T, Custom);
       setOperationAction(ISD::INSERT_VECTOR_ELT, T, Custom);
     }
-    if (EnableUnimplementedWasmSIMDInstrs) {
+    if (Subtarget->hasUnimplementedSIMD128()) {
       for (auto T : {MVT::v2i64, MVT::v2f64}) {
         setOperationAction(ISD::EXTRACT_VECTOR_ELT, T, Custom);
         setOperationAction(ISD::INSERT_VECTOR_ELT, T, Custom);
       }
     }
   }
 
   // Trap lowers to wasm unreachable
case ISD::CopyToReg:
     return LowerCopyToReg(Op, DAG);
   case ISD::INTRINSIC_WO_CHAIN:
     return LowerINTRINSIC_WO_CHAIN(Op, DAG);
   case ISD::EXTRACT_VECTOR_ELT:
   case ISD::INSERT_VECTOR_ELT:
     return LowerAccessVectorElement(Op, DAG);
   case ISD::INTRINSIC_VOID:
     return LowerINTRINSIC_VOID(Op, DAG);
+  case ISD::SIGN_EXTEND_INREG:
+    return LowerSIGN_EXTEND_INREG(Op, DAG);
   case ISD::VECTOR_SHUFFLE:
     return LowerVECTOR_SHUFFLE(Op, DAG);
   case ISD::SHL:
   case ISD::SRA:
   case ISD::SRL:
     return LowerShift(Op, DAG);
   }
 }
default:
       llvm_unreachable("Invalid tag!");
     }
     break;
   }
   }
 }
 
 SDValue
+WebAssemblyTargetLowering::LowerSIGN_EXTEND_INREG(SDValue Op,
+                                                  SelectionDAG &DAG) const {
+  // If sign extension operations are disabled, allow sext_inreg only if operand
+  // is a vector extract. SIMD does not depend on sign extension operations, but
+  // allowing sext_inreg in this context lets us have simple patterns to select
+  // extract_lane_s instructions. Expanding sext_inreg everywhere would be
+  // simpler in this file, but would necessitate large and brittle patterns to
+  // undo the expansion and select extract_lane_s instructions.
+  assert(!Subtarget->hasSignExt() && Subtarget->hasSIMD128());
+  if (Op.getOperand(0).getOpcode() == ISD::EXTRACT_VECTOR_ELT)
+    return Op;
+  // Otherwise expand
+  return SDValue();
+}
+
+SDValue
 WebAssemblyTargetLowering::LowerVECTOR_SHUFFLE(SDValue Op,
                                                SelectionDAG &DAG) const {
   SDLoc DL(Op);
   ArrayRef<int> Mask = cast<ShuffleVectorSDNode>(Op.getNode())->getMask();
   MVT VecType = Op.getOperand(0).getSimpleValueType();
   assert(VecType.is128BitVector() && "Unexpected shuffle vector type");
   size_t LaneBytes = VecType.getVectorElementType().getSizeInBits() / 8;
 

lib/Target/WebAssembly/WebAssemblyInstrInfo.td

 //===----------------------------------------------------------------------===//
 // WebAssembly Instruction Predicate Definitions.
 //===----------------------------------------------------------------------===//
 
 def HasAddr32 : Predicate<"!Subtarget->hasAddr64()">;
 def HasAddr64 : Predicate<"Subtarget->hasAddr64()">;
 def HasSIMD128 : Predicate<"Subtarget->hasSIMD128()">,
                            AssemblerPredicate<"FeatureSIMD128", "simd128">;
-def HasUnimplementedSIMD : Predicate<"EnableUnimplementedWasmSIMDInstrs">;
+def HasUnimplementedSIMD :
+    Predicate<"Subtarget->hasUnimplementedSIMD128()">,
+    AssemblerPredicate<"FeatureUnimplementedSIMD128", "unimplemented-simd128">;
 def HasAtomics : Predicate<"Subtarget->hasAtomics()">,
                            AssemblerPredicate<"FeatureAtomics", "atomics">;
 def HasNontrappingFPToInt :
     Predicate<"Subtarget->hasNontrappingFPToInt()">,
               AssemblerPredicate<"FeatureNontrappingFPToInt",
                                  "nontrapping-fptoint">;
 def NotHasNontrappingFPToInt :
     Predicate<"!Subtarget->hasNontrappingFPToInt()">,

lib/Target/WebAssembly/WebAssemblySubtarget.h

 #include <string>
 
 #define GET_SUBTARGETINFO_HEADER
 #include "WebAssemblyGenSubtargetInfo.inc"
 
 namespace llvm {
 
 class WebAssemblySubtarget final : public WebAssemblyGenSubtargetInfo {
-  bool HasSIMD128;
-  bool HasAtomics;
-  bool HasNontrappingFPToInt;
-  bool HasSignExt;
-  bool HasExceptionHandling;
+  enum SIMDEnum {
+    NoSIMD,
+    SIMD128,
+    UnimplementedSIMD128,
+  } SIMDLevel = NoSIMD;
+
+  bool HasAtomics = false;
+  bool HasNontrappingFPToInt = false;
+  bool HasSignExt = false;
+  bool HasExceptionHandling = false;
 
   /// String name of used CPU.
   std::string CPUString;
 
   /// What processor and OS we're targeting.
   Triple TargetTriple;
 
   WebAssemblyFrameLowering FrameLowering;
const WebAssemblyRegisterInfo *getRegisterInfo() const override {
     return &getInstrInfo()->getRegisterInfo();
   }
   const Triple &getTargetTriple() const { return TargetTriple; }
   bool enableMachineScheduler() const override;
   bool useAA() const override;
 
   // Predicates used by WebAssemblyInstrInfo.td.
   bool hasAddr64() const { return TargetTriple.isArch64Bit(); }
-  bool hasSIMD128() const { return HasSIMD128; }
+  bool hasSIMD128() const { return SIMDLevel >= SIMD128; }
+  bool hasUnimplementedSIMD128() const {
+    return SIMDLevel >= UnimplementedSIMD128;
+  }
   bool hasAtomics() const { return HasAtomics; }
   bool hasNontrappingFPToInt() const { return HasNontrappingFPToInt; }
   bool hasSignExt() const { return HasSignExt; }
   bool hasExceptionHandling() const { return HasExceptionHandling; }
 
   /// Parses features string setting specified subtarget options. Definition of
   /// function is auto generated by tblgen.
   void ParseSubtargetFeatures(StringRef CPU, StringRef FS);
 };
 
 } // end namespace llvm
 
 #endif

lib/Target/WebAssembly/WebAssemblySubtarget.cpp

WebAssemblySubtarget::initializeSubtargetDependencies(StringRef FS) {
   ParseSubtargetFeatures(CPUString, FS);
   return *this;
 }
 
 WebAssemblySubtarget::WebAssemblySubtarget(const Triple &TT,
                                            const std::string &CPU,
                                            const std::string &FS,
                                            const TargetMachine &TM)
-    : WebAssemblyGenSubtargetInfo(TT, CPU, FS), HasSIMD128(false),
-      HasAtomics(false), HasNontrappingFPToInt(false), HasSignExt(false),
-      HasExceptionHandling(false), CPUString(CPU), TargetTriple(TT),
-      FrameLowering(), InstrInfo(initializeSubtargetDependencies(FS)), TSInfo(),
+    : WebAssemblyGenSubtargetInfo(TT, CPU, FS), CPUString(CPU),
+      TargetTriple(TT), FrameLowering(),
+      InstrInfo(initializeSubtargetDependencies(FS)), TSInfo(),
       TLInfo(TM, *this) {}
 
 bool WebAssemblySubtarget::enableMachineScheduler() const {
   // Disable the MachineScheduler for now. Even with ShouldTrackPressure set and
   // enableMachineSchedDefaultSched overridden, it appears to have an overall
   // negative effect for the kinds of register optimizations we're doing.
   return false;
 }
 
 bool WebAssemblySubtarget::useAA() const { return true; }

test/CodeGen/WebAssembly/simd-arith.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128,SIMD128-SLOW
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,SIMD128,SIMD128-FAST
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128,SIMD128-SLOW
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,SIMD128,SIMD128-FAST
 ; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
 ; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-simd128 | FileCheck %s --check-prefixes CHECK,NO-SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -fast-isel | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; check that a non-test run (including explicit locals pass) at least finishes
-; RUN: llc < %s -O0 -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext
-; RUN: llc < %s -O2 -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext
+; RUN: llc < %s -O0 -mattr=+unimplemented-simd128
+; RUN: llc < %s -O2 -mattr=+unimplemented-simd128
 
 ; Test that basic SIMD128 arithmetic operations assemble as expected.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; ==============================================================================
 ; 16 x i8
define <16 x i8> @shr_s_v16i8(<16 x i8> %v, i8 %x) {
   %a = ashr <16 x i8> %v, %s
   ret <16 x i8> %a
 }
 
 ; CHECK-LABEL: shr_s_vec_v16i8:
 ; NO-SIMD128-NOT: i8x16
 ; SIMD128-NEXT: .functype shr_s_vec_v16i8 (v128, v128) -> (v128){{$}}
 ; SIMD128-NEXT: i8x16.extract_lane_s $push[[L0:[0-9]+]]=, $0, 0{{$}}
-; SIMD128-NEXT: i32.const $push[[L1:[0-9]+]]=, 24{{$}}
-; SIMD128-NEXT: i32.shl $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
-; SIMD128-NEXT: i32.const $push[[L3:[0-9]+]]=, 24{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]]{{$}}
-; SIMD128-NEXT: i8x16.extract_lane_u $push[[L5:[0-9]+]]=, $1, 0{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L6:[0-9]+]]=, $pop[[L4]], $pop[[L5]]{{$}}
-; SIMD128-NEXT: i8x16.splat $push[[L7:[0-9]+]]=, $pop[[L6]]{{$}}
+; SIMD128-NEXT: i8x16.extract_lane_u $push[[L1:[0-9]+]]=, $1, 0{{$}}
+; SIMD128-NEXT: i32.shr_s $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
+; SIMD128-NEXT: i8x16.splat $push[[L3:[0-9]+]]=, $pop[[L2]]{{$}}
 ; Skip 14 lanes
-; SIMD128:      i8x16.extract_lane_s $push[[L7:[0-9]+]]=, $0, 15{{$}}
-; SIMD128-NEXT: i32.const $push[[L8:[0-9]+]]=, 24{{$}}
-; SIMD128-NEXT: i32.shl $push[[L9:[0-9]+]]=, $pop[[L7]], $pop[[L8]]{{$}}
-; SIMD128-NEXT: i32.const $push[[L10:[0-9]+]]=, 24{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L11:[0-9]+]]=, $pop[[L9]], $pop[[L10]]{{$}}
-; SIMD128-NEXT: i8x16.extract_lane_u $push[[L12:[0-9]+]]=, $1, 15{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L13:[0-9]+]]=, $pop[[L11]], $pop[[L12]]{{$}}
-; SIMD128-NEXT: i8x16.replace_lane $push[[R:[0-9]+]]=, $pop[[L14:[0-9]+]], 15, $pop[[L13]]{{$}}
+; SIMD128:      i8x16.extract_lane_s $push[[L0:[0-9]+]]=, $0, 15{{$}}
+; SIMD128-NEXT: i8x16.extract_lane_u $push[[L1:[0-9]+]]=, $1, 15{{$}}
+; SIMD128-NEXT: i32.shr_s $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
+; SIMD128-NEXT: i8x16.replace_lane $push[[R:[0-9]+]]=, $pop{{[0-9]+}}, 15, $pop[[L2]]{{$}}
 ; SIMD128-NEXT: return $pop[[R]]{{$}}
 define <16 x i8> @shr_s_vec_v16i8(<16 x i8> %v, <16 x i8> %x) {
   %a = ashr <16 x i8> %v, %x
   ret <16 x i8> %a
 }
 
 ; CHECK-LABEL: shr_u_v16i8:
 ; NO-SIMD128-NOT: i8x16
define <8 x i16> @shr_s_v8i16(<8 x i16> %v, i16 %x) {
   %a = ashr <8 x i16> %v, %s
   ret <8 x i16> %a
 }
 
 ; CHECK-LABEL: shr_s_vec_v8i16:
 ; NO-SIMD128-NOT: i16x8
 ; SIMD128-NEXT: .functype shr_s_vec_v8i16 (v128, v128) -> (v128){{$}}
 ; SIMD128-NEXT: i16x8.extract_lane_s $push[[L0:[0-9]+]]=, $0, 0{{$}}
-; SIMD128-NEXT: i32.const $push[[L1:[0-9]+]]=, 16{{$}}
-; SIMD128-NEXT: i32.shl $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
-; SIMD128-NEXT: i32.const $push[[L3:[0-9]+]]=, 16{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]]{{$}}
-; SIMD128-NEXT: i16x8.extract_lane_u $push[[L5:[0-9]+]]=, $1, 0{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L6:[0-9]+]]=, $pop[[L4]], $pop[[L5]]{{$}}
-; SIMD128-NEXT: i16x8.splat $push[[L7:[0-9]+]]=, $pop[[L6]]{{$}}
+; SIMD128-NEXT: i16x8.extract_lane_u $push[[L1:[0-9]+]]=, $1, 0{{$}}
+; SIMD128-NEXT: i32.shr_s $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
+; SIMD128-NEXT: i16x8.splat $push[[L3:[0-9]+]]=, $pop[[L2]]{{$}}
 ; Skip 6 lanes
-; SIMD128:      i16x8.extract_lane_s $push[[L7:[0-9]+]]=, $0, 7{{$}}
-; SIMD128-NEXT: i32.const $push[[L8:[0-9]+]]=, 16{{$}}
-; SIMD128-NEXT: i32.shl $push[[L9:[0-9]+]]=, $pop[[L7]], $pop[[L8]]{{$}}
-; SIMD128-NEXT: i32.const $push[[L10:[0-9]+]]=, 16{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L11:[0-9]+]]=, $pop[[L9]], $pop[[L10]]{{$}}
-; SIMD128-NEXT: i16x8.extract_lane_u $push[[L12:[0-9]+]]=, $1, 7{{$}}
-; SIMD128-NEXT: i32.shr_s $push[[L13:[0-9]+]]=, $pop[[L11]], $pop[[L12]]{{$}}
-; SIMD128-NEXT: i16x8.replace_lane $push[[R:[0-9]+]]=, $pop[[L14:[0-9]+]], 7, $pop[[L13]]{{$}}
+; SIMD128:      i16x8.extract_lane_s $push[[L0:[0-9]+]]=, $0, 7{{$}}
+; SIMD128-NEXT: i16x8.extract_lane_u $push[[L1:[0-9]+]]=, $1, 7{{$}}
+; SIMD128-NEXT: i32.shr_s $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]{{$}}
+; SIMD128-NEXT: i16x8.replace_lane $push[[R:[0-9]+]]=, $pop{{[0-9]+}}, 7, $pop[[L2]]{{$}}
 ; SIMD128-NEXT: return $pop[[R]]{{$}}
 define <8 x i16> @shr_s_vec_v8i16(<8 x i16> %v, <8 x i16> %x) {
   %a = ashr <8 x i16> %v, %x
   ret <8 x i16> %a
 }
 
 ; CHECK-LABEL: shr_u_v8i16:
 ; NO-SIMD128-NOT: i16x8

test/CodeGen/WebAssembly/simd-bitcasts.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test that bitcasts between vector types are lowered to zero instructions
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; CHECK-LABEL: v16i8_to_v16i8:
 ; NO-SIMD128-NOT: return $0

test/CodeGen/WebAssembly/simd-comparisons.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test SIMD comparison operators
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; CHECK-LABEL: compare_eq_v16i8:
 ; NO-SIMD128-NOT: i8x16

test/CodeGen/WebAssembly/simd-conversions.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test that vector float-to-int and int-to-float instructions lower correctly
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; CHECK-LABEL: convert_s_v4f32:
 ; NO-SIMD128-NOT: i32x4

test/CodeGen/WebAssembly/simd-ext-load-trunc-store.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 | FileCheck %s
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 | FileCheck %s
 
 ; Check that store in memory with smaller lanes are loaded and stored
 ; as expected. This is a regression test for part of bug 39275.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; CHECK-LABEL: load_ext_2xi32:

test/CodeGen/WebAssembly/simd-intrinsics.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,SIMD128
 
 ; Test that SIMD128 intrinsics lower as expected. These intrinsics are
 ; only expected to lower successfully if the simd128 attribute is
 ; enabled and legal types are used.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 

test/CodeGen/WebAssembly/simd-load-store-alignment.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128 | FileCheck %s
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 | FileCheck %s
 
 ; Test loads and stores with custom alignment values.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; ==============================================================================
 ; 16 x i8

test/CodeGen/WebAssembly/simd-offset.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-keep-registers -wasm-disable-explicit-locals | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test SIMD loads and stores
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; ==============================================================================
 ; 16 x i8

test/CodeGen/WebAssembly/simd-select.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128 | FileCheck %s
 
 ; Test that vector selects of various varieties lower correctly to bitselects.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; ==============================================================================
 ; 16 x i8

test/CodeGen/WebAssembly/simd-sext-inreg.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+unimplemented-simd128 | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+simd128 | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -wasm-keep-registers -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test that vector sign extensions lower to shifts
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; CHECK-LABEL: sext_inreg_v16i8:
 ; NO-SIMD128-NOT: i8x16

test/CodeGen/WebAssembly/simd.ll

-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+unimplemented-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128
 ; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,SIMD128-VM
-; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-simd128,+sign-ext | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers | FileCheck %s --check-prefixes CHECK,NO-SIMD128
 
 ; Test that basic SIMD128 vector manipulation operations assemble as expected.
 
 target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
 target triple = "wasm32-unknown-unknown"
 
 ; ==============================================================================
 ; 16 x i8

test/MC/WebAssembly/basic-assembly.s

-# RUN: llvm-mc -triple=wasm32-unknown-unknown -mattr=+simd128,+nontrapping-fptoint,+exception-handling < %s | FileCheck %s
+# RUN: llvm-mc -triple=wasm32-unknown-unknown -mattr=+unimplemented-simd128,+nontrapping-fptoint,+exception-handling < %s | FileCheck %s
 # this one is just here to see if it converts to .o without errors, but doesn't check any output:
-# RUN: llvm-mc -triple=wasm32-unknown-unknown -filetype=obj -mattr=+simd128,+nontrapping-fptoint,+exception-handling < %s
+# RUN: llvm-mc -triple=wasm32-unknown-unknown -filetype=obj -mattr=+unimplemented-simd128,+nontrapping-fptoint,+exception-handling < %s
 
     .text
     .section .text.main,"",@
     .type    test0,@function
 test0:
     # Test all types:
     .functype   test0 (i32, i64) -> (i32)
     .eventtype  __cpp_exception i32

test/MC/WebAssembly/simd-encodings.s

-# RUN: llvm-mc -show-encoding -triple=wasm32-unkown-unknown -mattr=+sign-ext,+simd128 < %s | FileCheck %s
+# RUN: llvm-mc -show-encoding -triple=wasm32-unkown-unknown -mattr=+unimplemented-simd128 < %s | FileCheck %s
 
 main:
     .functype main () -> ()
 
     # CHECK: v128.load 48:p2align=0 # encoding: [0xfd,0x00,0x00,0x30]
     v128.load 48
 
     # CHECK: v128.store 48:p2align=0 # encoding: [0xfd,0x01,0x00,0x30]

test/MC/WebAssembly/types.ll

-; RUN: llc -wasm-enable-unimplemented-simd -mattr=+sign-ext,+simd128 -filetype=obj %s -o - | obj2yaml | FileCheck %s
+; RUN: llc -mattr=+unimplemented-simd128 -filetype=obj %s -o - | obj2yaml | FileCheck %s
 
 target triple = "wasm32-unknown-unknown"
 
 declare i32 @i32()
 declare i64 @i64()
 declare float @f32()
 declare double @f64()
 declare <16 x i8> @v16i8()